Bacterial resistance to antibiotics has raised fears of an approaching medical catastrophe (Neu, Science, 257, 1064-1073 (1992)). Evolutionary selection and genetic transformation have made this problem pressing. Most antibiotic drugs are derivatives of naturally occurring bactericides (Davies, Science, 264, 375-382 (1994)), and many resistance mechanisms evolved long ago. Human use of antibiotics has refined these mechanisms and promoted their spread through gene transfer (Davies, Science, 264, 375-382 (1994)). A resistance mechanism originating in one species of bacteria can be expected to spread throughout the biosphere.
Bacterial adaptations to .beta.-lactam drugs (e.g., amoxicillin, cephalothin, clavulanate, aztreonam) are among the best studied and most pernicious forms of antibiotic resistance. .beta.-lactams target enzymes that are unique to bacteria and are thus highly selective. They have been widely prescribed. In the absence of resistance, .beta.-lactams are the first choice for treatment in 45 of 78 common bacterial infections (Goodman & Gilman's The Pharmacological Basis of Therapeutics (Hardman et al., eds., McGraw-Hill, New York, 1996)). The evolution of resistance to these drugs has raised the cost of antibiotic therapy and reduced its effectiveness, leading to increased rates of morbidity and mortality.
.beta.-lactam antibiotics inhibit bacterial cell wall biosynthesis (Tomasz, Rev. Infect. Dis., 8, S270-S278 (1986)). The drugs form covalent complexes with a group of transpeptidases/carboxypeptidases called penicillin binding proteins (PBPs). PBP inactivation disrupts cell wall biosynthesis, leading to self-lysis and death of the bacteria.
Bacteria use several different mechanisms to escape from .beta.-lactam drugs (Sanders, Clinical Infectious Disease, 14, 1089-1099 (1992); Li et al., Antimicrob. Agents Chemother., 39, 1948-1953 (1995)). Probably the most widespread is the hydrolysis of .beta.-lactams by .beta.-lactamase enzymes.
TEM-1 and AmpC are two .beta.-lactamases from Escherichia coli. E. coli is an important pathogen in its own right. It is the most common cause of gram-negative bacterial infection in humans (Levine, New Engl. J. Med., 313, 445-447 (1985)), and is the most prevalent hospital-acquired infection (Thornsberry, Pharmacotherapy, 15, S3-8 (1995)). E. coli that carry TEM-1, or for which AmpC production has been derepressed, are resistant to .beta.-lactam treatment. As of 1992, as many of 30% of community-isolated E. coli and 40-50% of hospital-acquired E. coli in the United States were resistant to .beta.-lactams such as amoxicillin (Neu, Science, 257, 1064-1073 (1992)). Many of these resistant E. coli are resistant to .beta.-lactamase inhibitors such as clavulanic acid and sulbactam.
TEM-1 and AmpC are major forms of plasmid-based and chromosomal .beta.-lactamases and are responsible for resistance in a broad host range. The versions of TEM and AmpC (Galleni, et al., Biochem. J., 250, 753-760 (1988)) in other bacterial species share high sequence identity to TEM-1 and AmpC from E. coli. TEM-1 structurally and catalytically resembles the class A .beta.-lactamase from Staphlococcus aureus. The structures of AmpC from Citrobacter freundii and Enterobacter cloacae have been determined, and they closely resemble the structure of the E. coli enzyme (K. Usher, L. Blaszczak, B. K. Shoichet, J. R. Remington, in preparation (1996)).
To overcome the action of .beta.-lactamases, medicinal chemists have introduced compounds that inhibit these enzymes, such as clavulanic acid, or compounds that are less susceptible to enzyme hydrolysis, such as aztreonam. Both have been widely used in antibiotic therapy (Rolinson, Rev. Infect. Diseases 13, S727-732 (1991)); both are .beta.-lactams. Their similarity to the drugs that they are meant to protect or replace has allowed bacteria to evolve further, maintaining their resistance.
Resistance to these new classes of .beta.-lactams has arisen through modifications of previously successful mechanisms. Point substitutions in .beta.-lactamases allow the enzymes to hydrolyze compounds designed to evade them (Philippon et al., Antimicrob. Agents Chemother., 33, 1131-1136 (1989)). Other substitutions reduce the affinity of .beta.-lactam inhibitors for the enzymes (Saves, et al., J. Biol. Chem., 270, 18240-18245 (1995)) or allow the enzymes to simply hydrolyze them. Several gram positive bacteria, such as Staph. aureus, have acquired sensor proteins that detect .beta.-lactams in the environment of the cell (Bennet and Chopra, Antimicrob. Agents Chemotherapy, 37, 153-158 (1993)). .beta.-lactam binding to these sensors leads to transcriptional up-regulation of the .beta.-lactamase. .beta.-lactam inhibitors of .beta.-lactamases, thus, can induce the production of the enzyme that they are meant to inhibit, defeating themselves.
It is noteworthy that the human therapeutic attack on bacteria has paralleled the path taken in nature. Several species of soil bacteria and fungi produce .beta.-lactams, presumably as weapons against other bacteria (although this remains a matter of debate). Over evolutionary time, susceptible bacteria have responded to .beta.-lactams with .beta.-lactamases, among other defenses. In turn, soil bacteria have produced .beta.-lactams that resist hydrolysis by .beta.-lactamases or have produced .beta.-lactams that inhibit the .beta.-lactamases. Streptomyces clavuligeris makes several .beta.-lactams, including clavulanic acid, a clinically used inhibitor of class A .beta.-lactamases such as TEM-1. Chromobacterium violaceum makes aztreonam, a clinically used monobactam that resists hydrolysis by many .beta.-lactamases. One reason why bacteria have been able to respond rapidly with "new" resistance mechanisms to .beta.-lactams, and indeed many classes of antibiotics, is that the mechanisms are not in fact new. As long as medicinal chemistry focuses on new .beta.-lactam molecules to overcome .beta.-lactamases, resistance can be expected to follow shortly. The logic will hold for any family of antibiotic where the lead drug, and resistance mechanisms to it, originated in the biosphere long before their human therapeutic use. This includes the aminoglycosides, chloramphenicol, the tetracyclines and vancomycin.
One way to avoid recapitulating this ancient "arms race" would be to develop inhibitors that have novel chemistries, dissimilar to .beta.-lactams. These non-.beta.-lactam inhibitors would not themselves be degraded by .beta.-lactamases, and mutations in the enzymes should not render them labile to hydrolysis. Novel inhibitors would escape detection by .beta.-lactam sensor proteins that up-regulate .beta.-lactamase transcription, and may be unaffected by porin mutations that limit the access of .beta.-lactams to PBPs. Such inhibitors would allow current .beta.-lactam drugs to work against bacteria where .beta.-lactamases provide the dominant resistance mechansim.
It has previously been reported that boric acid and certain phenyl boronic acids are inhibitors of certain .beta.-lactamases. See, Kiener and Waley, Biochem. J., 169, 197-204 (1978) (boric acid, phenylboronic acid (2FDB) and m-aminophenylboronate (MAPB)); Beesley et al., Biochem. J., 209, 229-233 (1983) (twelve substituted phenylborinic acids, including 2-formylphenylboronate (2FORMB), 4-formylphenylboronate (4FORMB), and 4-methylphenylboronate (4MEPB)); Amicosante et al., J. Chemotherapy, 1, 394-398 (1989) (boric acid, 2FDB, MAPB and tetraphenylboronic acid). More recently, m-(dansylamidophenyl)-boronic acid (NSULFB) has been reported to be a submicromolar inhibitor of the Enterobacter cloacae P99 .beta.-lactamase. Dryjanski and Pratt, Biochemistry, 34, 3561-3568 (1995). In addition, Strynadka and colleagues used the crystallographic structure of a mutant TEM-1 enzyme-penicillin G complex to design a novel alkylboronic acid inhibitor [(1R)-1-acetamido-2-(3-carboxyphenyl)ethane boronic acid] with high affinity for this enzyme. Strynadka et al., Nat. Struc. Biol., 3, 688-695 (1996).